They used to be called GAFA, now I understand they are the Intel 7*. Whatever. These are the folks who are going to disrupt your business. Or so goes the narrative. It is a little hard to see how these virtual (as opposed to bricks and mortar) businesses will really disrupt oil and/or gas, although I suppose such a case could be made for disruption in trading. This is about as likely to happen as GAFA disrupting the banking business which is proving a lot harder than originally thought. But I digress. The ‘pure’ i.e. Silicon-Valley style disruption play is unlikely to work in a diversified, established and technologically challenging field like oil and gas. So the narrative has been spun around the impact of the information technology that is spinning out of the GAFAs and into business, a phenomenon known as ‘digital transformation.’
The narrative is wrong on two counts. First the oil and gas industry is an extremely computer-literate business with a large installed base, decades of R&D and a very recent bout of ‘disruption’ in the form of the ‘digital oilfield’ movement of the 2000s. Oil and gas is not your yellow/black taxi cab business waiting around to be disrupted by an Uber. The other reason the current narrative is wrong is that none of the GAFAs are really IT businesses. They are combinations of a great idea for a new business or the result of ruthless exploitation of somebody else’s great idea, followed by a first mover/monopolistic/winner-takes-all situation. Despite appearances, the compute/IT side is totally subordinate to the business. If you doubt this, think of how freely Google gives away its IT and contrast this with how jealously it guards its business data!
The disruption narrative, although false, has been successful in the IT/consulting community which has is always on the lookout for a new way of creating fear, uncertainty and doubt. The IT community has grabbed the wrong end of the stick in Google’s success (an awkward metaphor, sorry) and is rushing around trying to convince all and sundry (perhaps that should be oil and sundry?) that they need Hadoop, BigTable Spark and what have you, all the shiny new stuff that has ‘enabled’ the disruption elsewhere.
There is another weakness in this software free-for-all that derives originally from the GAFAs in-house requirements. It is worthwhile reflecting on exactly what these are. I’m not an expert on Google’s IT, but just, for a minute, imagine what the internet looks like from Google HQ. It is the mother of all data deluges! Bazillions of clickstreams coming in from internauts all over the world in real time. Somehow all this needs capturing (more on this in ‘the future of seismic data storage’ on page 8) and turning into a database of who clicked on what, where and when.
In one sense, this is mindbogglingly smart. In another it is stultifying dumb. For an oil, an immediate application of the Hadoop ecosystem is in extracting business-relevant information from computer log files, for instance in the cyber security field. This rather dumb application maps pretty well to the inverted clickstream model.
Needless to say, those who are selling disruption have chosen to emphasize the mindboggling. Hence the talk of scalable systems, big data and artificial intelligence. And, more generally, on solving problems which you did not even realized you had! The Hadoop ecosystem (whose main job is to do the dumb stuff) has been extended with an Alphabet (sic) soup of acronyms that point us in the direction of the smart stuff. I, like everyone else, am much more interested in the smarts, even if I doubt that all of it is terribly relevant. Much of the big data/artificial intelligence tools are ‘highly scalable’ hammers looking for nails. I sometimes wonder if Google didn’t chuck all this stuff over the perimeter wall as a diversion. Rather like the British are said to have done in WWII by spreading the rumor that pilots were given carrots to help them see in the dark, when in fact, Radar has just been invented.
One company at the cutting edge of all this is GE whose departing CEO Jeff Immelt, was one of the first to put his money and mouth on big data and the Industrial Internet of Things. Five years ago, Immelt stated that ‘The industrial internet is revolutionizing our services. We will leverage our $150 billion backlog to grow this technology and our revenues 4-5% annually.’ A year later Jesse Demesa revealed GE had sunk $1 billion into the IIoT and that around 1,000 developers were beavering away on Predix, GE’s core industrial platform. Predix by the way is quite a remarkable infrastructure in that it shares much of the GAFA-derived open source software stack and also, GE really has been pretty ‘open’ itself in showing Predix’ innards to the world.
Five years is a long time at ‘internet speed’ and you might expect some signs of success. This does not appear to be materializing. The Financial Times of 21st October performs quite a hatchet job on Immelt and on GE. The FT’s Lex column has it that ‘Immelt’s abrupt departure suddenly makes sense [as the] scale of the debacle he created [..] became clear.’ Immelt’s replacement John Flannery spoke of the need to make ‘major changes with urgency.’ The FT editorial has it that GE’s bets on the IIoT ‘have yet to pay off.’ There is talk of sell-offs. One might think that Predix would be sell-off opportunity. But IMHO this is a case where the part would be worth a lot less than the whole. For GE and Google, the value is in the data not the software. That’s why Google gives its stuff away for free!
* Google, Amazon, Facebook, Apple. For the Intel 7 see below.
© Oil IT Journal - all rights reserved.